How does probability relate to machine learning?

Hello, I’m new to Kaggle and would like to work on the Bosch competition. My naive approach…assuming I didn’t know about scikit-learn…would be to look at all the product failures and which factors contributed to those failures, then calculate a failure rate probability for each factor. Then, I would start looking at combinations of factors and calculate the probability of two factors producing a failure. This would work for awhile…looking at more and more combinations…until it becomes computationally intractable. Then I would have to start screening by confidence levels and choose only those factors with the biggest impact. Basically…I think I am describing either random forest or factor analysis. I haven’t done ML yet, so I’m a bit clueless. I am good with both python and pandas (and mathematica). One…


Link to Full Article: How does probability relate to machine learning?